Yes, no matter what approach we take, 'the user still needs to know what does or will happen given a certain set of crossover parameters'.
Whether you use sims alone, measurements alone, or the combination of the two, all approaches require that.
Taking the requisite knowledge as a given for comparison's sake,
multi-way active with measurements can certainly eliminate the need for simulations, much more so than passive...imho.
Can a simulator tell you things about a cabinet that measurements or ears might struggle with, like diffraction issues, for example, particularly edge diffraction?
Last edited:
Yes I will workout the woofer box volume and tuning in a simulator usually use Bass Box Pro and then design the cabinet around the volume. I use 1/4 rounds and don't go into detail in the simulator about bracing and such.
Rob🙂
I think quite possibly, quite probably so. Others certainly know more here than i do.
But i kind of end up scratching my head over what can really be done about diffraction, other than maybe a few common sense things like roundovers, etc.
I've tried running measurements as I attach different roundovers, felt strips, and such, ....which yes, sometimes helps the measurements.
But not enough yet, to convince me it's worth much pursuit.
"You should give most credence to your ears", nope..."Human perception is rife with always getting it wrong...that's why we have machines" Neil deGrasse Tyson. Be it vision, hearing, touch, all of it...numbers don't lie, mathematics are absolute, 2+2 equals four thru-out our universe. Our perceptions are at fault...
--------------------------------------------------------------------------Rick........
--------------------------------------------------------------------------Rick........
I could easily agree that "xover simulation" is not the best term for process where we control acoustical domain with calculated electrical transfer functions. Differences exist when "make adjustments" happen.
Simulation process calculates acoustical domain in radiation space while adjusting or calculating electrical transfer functions so measurements and adjustment with complete speaker are not necessarily needed. For example I don't normally measure and tune complete speaker designed with simulator able to visualize radiation space in acoustical domain before loading parameters to gear or app. finally realizing electrical transfer functions.
The biggest problem for common DIY enthusiast is that he/she cannot measure free field responses of complete speaker for adjustments. Simulation of whole construction in acoustical domain is at least possible with some uncertainties with measurement data, but also much faster and reliable compared measurements in reverberation chamber called as living room.
Just your style, which does not eliminate xover simulation from anybody else. The least from myself 🙂
Last edited:
Sorry if i didn't communicate this well, that if starting proto measurements and sims are any where reasonable with each other, I'm then done with further prototype sims.
So yes, I totally have more faith in measurements...1000% more.
Honestly I don't see how anyone could not trust well made and repeatable measurements more than sims..??
All that makes good sense.
And I'd like to thank you for your generosity towards facilitating simulations.
I learn a great deal from following the guys who know how to use VCad.
Heck, if i hadn't spent so much time learning to measure, and if BB plywood weren't so cheap here, I'd probably be learning to sim a heck of alot more too!
And yet people seem not to. Perhaps it's not that easy to do well made measurements. How many measurements do you need to take to produce a picture of what's going on as thoroughly as a sim does? There's no reason I see why a sim shouldn't be accurate so long as the input is accurate.
Good observations, good questions.
It's true for sure, making good measurements is quite difficult to learn and implement.
And if i may quote kimmosto from post #26 "The biggest problem for common DIY enthusiast is that he/she cannot measure free field responses of complete speaker for adjustments. Simulation of whole construction in acoustical domain is at least possible with some uncertainties with measurement data, but also much faster and reliable compared measurements in reverberation chamber called as living room" .
I think these two factors are a couple of big reasons why simulations enjoy popularity....and rightly so.
As to, how many measurements does it take to get a picture of what's going on, as thoroughly as with a sim...
Well, i'd like to redirect the question and ask, how many sims does it take to match the most simple direct on axis measurement, which i can't deny?
My experience so far, has been a bunch.
So how much credence can i put in all the info the sims purport to tell?
As to, no reason why sims shouldn't be accurate...
Can totally agree, if all the variables, and most importantly the underlying assumptions the simulator uses, are in sync with realistic builds.
What i'd love to know, is how many folks think their (well made and repeatable) measurements ended up matching their sims?
My sims, even on very simple builds, have been close but no cigar.
I usually use sims to see whether a drive unit / box / horn will get close to what I need or not. Once I've decided on drive units I build test boxes / horns and measure to see what I have in the real world. I then sort out my xo on real measurements not computer guesstimates.
My ears can tell if there's harshness / problems in a driver / box combo, but are not good enough to tell me the exact frequency range that is causing it. My measurement mic helps a lot in finding out what's causing problems.
One thing I do quite a lot is jump between headphones and speakers to get the tonal balance right in room.
Rob.
My ears can tell if there's harshness / problems in a driver / box combo, but are not good enough to tell me the exact frequency range that is causing it. My measurement mic helps a lot in finding out what's causing problems.
One thing I do quite a lot is jump between headphones and speakers to get the tonal balance right in room.
Rob.
To me that is the wrong question to ask of a simulator.
BEM simulations can give good insight into directivity, diffraction, cabinet shape influence etc. All without having to make any sawdust. They can even be free of cost other than time (for non commercial use of ABEC).
Vituix can do all sorts of stuff but being able to see the impact of crossover / EQ changes to all off axis angles without having to remeasure is appealing. Take one set of really good data and run with it.
You are not like most diyers in many ways and so your viewpoint is different. You have the equipment, skill and venue to take good almost anechoic measurements. You can also prototype things quickly and at a cost that is reasonable to you. For you that way makes the most sense no reason not to use it 😉
Hello Mark100
What do you consider close?? If mine are in a db or 2 of the simulation it works for me. Taking a post build measurement is the only way to confirm that you ended up with what you intended by design.
Here is a comparison of the last compensation build I did as an example.
Red is predicted Blue is measured.
Rob 🙂
Attachments
Last edited:
Simulations regarding filter changes and electrical matters are usually more successful than some other practical sims currently are.
Hello
What type of simulations are you talking about? I use a box simulation and a crossover simulation. What others would you need?
Rob🙂
Mainly spice, also crossover electrical simulators, driver resonance in a box types as well are known to give fairly predictable results.
"Simulations, Measurements and Ears"
All three equally, examining and absorbing all three and making changes until a high level of satisfaction is reached!
All three equally, examining and absorbing all three and making changes until a high level of satisfaction is reached!
I believe our ears tell us more that what can be simulated. Quantifying diffraction is difficult. Compound curves and felt alter the FR shape, the unfiltered ripple response will change a little(micro) but audibly it appears there's a little more going on that reduces the earliest reflections and unveils the the recorded L/R and depth architecture.
Simulations, and the high quality measurements which feed the simulations, are absolutely necessary to get a speaker system to the 95% point. But the last 5% needs to be done by ear. Such things as setting tweeter levels up or down by 1 dB... Staring at graphs is not going to help me make that decision.
